Criminal Justice Section  


Criminal Justice Magazine
Fall 2001
Volume 16, Issue 3

The Eyes Have It—Or Do They?

New Guides for Better Eyewitness Evidence Procedures

By James M. Doyle, Mark R. Larson, and Caterina M. DiTraglia

Who hasn’t seen the sensational news reports of cases in which DNA evidence has been used to exonerate individuals previously convicted of crimes? What evidence, then, was used to obtain these convictions in the first place? And how could justice have repeatedly gone so awry?

When researchers and criminal justice practitioners began to ask these and corollary questions, they uncovered some startling results. Of the first 60 wrongful convictions revealed by DNA technology, 53 had relied to some extent on confident, but mistaken, eyewitnesses. (Dwyer, Sheck & Neufeld, Actual Innocence (2000).) Perhaps our oldest prosecutorial method—that of eyewitness accounts—was failing to prevent the two things that no one in the criminal justice system wants to see: the wrongful conviction of an innocent person and the consequent escape of a guilty one.

To address these questions and to suggest solutions, the National Institute of Justice (NIJ) convened the Technical Working Group for Eyewitness Evidence in 1998. This group was comprised of law enforcement, legal, and research professionals who joined forces and shared information with the goal of developing improved protocols for collecting and preserving eyewitness evidence so that the most accurate and reliable evidence could be presented to judges and juries. Members of the group came from rural and urban areas, from across America and Canada, with the hopes that they could recommend procedures that would be implemented and followed across the land. Yet, rather than suggest that one set of rules become mandated practice everywhere, the group recognized that various locales have unique conditions and methods that must be honored and adhered to. What they eventually agreed to was published in October 1999 under the title Eyewitness Evidence: A Guide for Law Enforcement. It represents a compilation of refined investigative practices that can be used by various jurisdictions as a tool to shape their local practices with regard to the handling of eyewitness evidence. It may also one day be seen as a landmark in the relationship between professionals in the criminal justice system and psychological researchers.

A long history of distrust

Despite the close cooperation that manifested itself in the production of A Guide for Law Enforcement, a deep-seated ambivalence has existed between legal and psychological professionals for quite some time. It may have gotten its start in 1909 when John Henry Wigmore, dean and professor of the law of evidence at Northwestern University, relentlessly attacked psychology professor, Hugo Munsterberg. ( See generally John Henry Wigmore, Professor Munsterberg and the Psychology of Testimony, 3 Ill. L. Rev. 399 (1909).) Munsterberg was then a respected applied psychologist who headed Harvard University’s psychology program and had expressed views regarding the fallibility of eyewitness evidence. Dean Wigmore carried his criticism further, remarking that when the psychologists were ready for the courts, the courts would be ready for them. John Brigham writes that Wigmore’s comment touched off "an often acrimonious debate between psychologists and the legal community which continued intermittently for the rest of the century." ( See John Brigham, What Is Forensic Psychology Anyway? 23 L. & Hum. Behav. 273, 274 (1999).)

Yet even as law and psychology have experienced difficulty in living together, they have also been unwilling to live entirely apart. The problem is that the places they inhabit, and their languages, seem so distant from one another. Psychologists conduct controlled experiments that yield statistical, probability- based results. Law enforcement officers and defense and prosecuting lawyers, on the other hand, need to know how to translate the psychologists’ statistical information into a forensic context that is essentially clinical and diagnostic. In the context of the use of psychological research findings about the fragility of eyewitness memory, lawyers are not concerned with the extent to which stress or other factors may degrade witness accuracy generally. Rather, courts must decide whether the specific case before them is one in which witness accuracy may have been adversely affected.

This fundamental difference in perspectives is aggravated by lawyers’ endemic tendency to see all questions of science as questions of expert testimony. Talk to defense lawyers or prosecutors about the psychology of eyewitness identification and you will find that neither group permits you to stray very far from the following topics: (1) How do I get an expert on the stand? (2) How do I keep their expert off the stand? and (3) How do I destroy their expert if they do get him or her on the stand? When lawyers can be persuaded to read the basic psychological materials at all, they do so because they are convinced that it will help them achieve one of those goals.

The partisanship of the American criminal justice system’s lawyers also confounds true understanding of the issues: one is either a defender or a prosecutor, and the same expert testimony that hurts you in today’s case will likely hurt you again tomorrow. One natural result of this has been to confine to a bare minimum the psychological knowledge that did flow into the criminal process.

Lawyers caught up in the expert witness debate often completely ignored the psychologists’ concentration on specific factors affecting eyewitness memory to engage in a tug-of-war over the level of general skepticism regarding eyewitness reliability. Defense lawyers argued that you shouldn’t believe eyewitnesses; prosecutors countered that you should. The former contended that you should believe all of the research pointing to the fallibility of eyewitness testimony; prosecutors asserted that you should believe none of it. This eyewitness expert controversy did have some collateral benefits, however. For example, it allowed defense lawyers to hold open a gap through which new psychological research, which was growing in quantity and improving in quality during the 1980s and 1990s, could find its way into the justice system. It also forced the researchers to undergo adversarial testing that was uncomfortable for individual expert witnesses, but helpful for the field as a whole. Nonetheless, the overall impact of the expert debate was debilitating. It came to rest at exactly the place where responsible psychologists would argue that it did not belong.

The regrettable upshot of this expert witness tug-of-war was that any useful discussion about system improvement became far more difficult. The parties that needed to make effective changes were staked out in opposing camps. As the battle lines were drawn, it seemed that we confronted a zero-sum choice between too many erroneous convictions and too many erroneous acquittals.

Impetus for change

Two forces slowly reshaped this landscape. First, the Supreme Court in Daubert v. Merrell-Dow Pharmaceuticals, Inc., 113 S. Ct. 2786 (1993), adopted as its definition of scientific evidence a version previously developed by the United States Court of Appeals for the Third Circuit in United States v. Downing, 753 F.2d 1224 (3d Cir. 1985), a case that happened to involve expert eyewitness testimony. Daubert made it more difficult for legal practitioners simply to shrug off psychological research as "bad science." Meanwhile, outside the courthouse, advances in DNA technology were demonstrating in a convincing and media-friendly fashion that innocent people were indeed being convicted on the basis of mistaken identification testimony. If there were any lingering doubts, Dwyer, Scheck, and Neufeld laid them to rest in their study, mentioned above, showing that more than 85 percent of the initial convictions overturned by DNA evidence had relied on eyewitness testimony.

Thus, a new opportunity for cooperation seemed to be presenting itself, one that harkened back to Wigmore’s remarks of 1909 when, after criticizing Munsterberg’s psychology, he nonetheless postulated a "friendly and energetic alliance of psychology and law in the noble cause of justice" in the future. ( See John Henry Wigmore, Professor Munsterberg and the Psychology of Testimony, 3 Ill. L. Rev. 399, 432 (1909).) A common ground was emerging so that a realistic and considered examination of the system seemed to be in order because there were principles to which both sides could agree.

First, all could grant that no one wants the conviction of innocent persons. Although some may debate the extent to which wrongful convictions are a problem, pragmatic law enforcement officials can see that a continuous drumbeat of publicity regarding DNA exonerations of wrongly convicted defendants will have the effect of creating a steep rise in general cynicism on the part of the public—and the juries drawn from the public—concerning eyewitness accuracy. It will become harder to get the convictions that you should get when the public is being barraged with the convictions you should not have gotten 10 years ago.

Second, a distinct body of research into "system variables" began to gain attention, ably advocated by Professor Gary Wells of Iowa State University. This research, which analyzed the impact of the legal system itself on eyewitnesses, has a number of features that made it more attractive to criminal justice system professionals. Professor Wells was performing experiments that suggested that the methods used in constructing and conducting lineups and photo spreads of criminal suspects were contributing to false witness identifications. His experiments were more palatable to law enforcement personnel because they seemed more "scientific"—i.e., they were easier to replicate, easier to explain, and more ecologically plausible than findings about "witness factors" such as stress or cross-racial identification. Moreover, the shift away from generalized findings about witness variables—which law enforcement personnel could do little or nothing to change—to the new research on system variables created an incentive for law enforcement to examine its own standards and procedures. Good police departments want to do things right. Here was convincing evidence that they could take logical steps that would improve the quality of eyewitness investigations. Furthermore, the studies showed no real downside. The researchers were suggesting changes that significantly reduced the number of false identifications without significantly lowering the number of correct identifications.

In short, law enforcement professionals realized that the psychologists had something to say, even if those officers of the law still believed that the psychologists should not be allowed to voice it on the witness stand. The key principle that law enforcement drew from the research was that eyewitness memory should be viewed as "trace evidence," subject to possible contamination, and requiring "rigorous criteria for handling eyewitness evidence that are as demanding as those governing the handling of physical trace evidence." ( See National Institute of Justice, Eyewitness Evidence: A Guide for Law Enforcement 2 (1999).)

The Guide for Law Enforcement experience

With the goal reconceived as using psychological insights to suggest ways that the police could get things right in the future, actual alliances, however temporary and wary, became a possibility. The NIJ seized the moment and called together 34 forensic psychological researchers, police investigators, defense lawyers, and prosecutors, and the Technical Working Group for Eyewitness Evidence was born. The product of their collaboration, Eyewitness Evidence: A Guide for Law Enforcement, illustrates both the progress that can be made in improving eyewitness investigations and the potential challenges to further progress.

The Guide takes several steps forward, especially because it looks closely at the basic elements of police investigations and suggests workable alterations to them to achieve more consistent eyewitness results. For example, standard police questioning practices have been designed more to elicit information from recalcitrant and antagonistic suspects than to foster accurate recall from witnesses who are trying to cooperate with enforcement officers. This can have a destructive impact on eyewitness memory by contaminating the witnesses’ memories with data unintentionally inserted by the officers, information that may prove completely unreliable. The "cognitive interview" techniques that are advocated by the researchers—including the rigorous avoidance of leading questions, a heightened awareness of social dynamics, and memory enhancement techniques such as encouraging witnesses to reinstate the context of the events—increase the amount of information gathered in eyewitness interviews and ensure that the information is supplied only by the eyewitness.

The Guide also includes an array of painstaking procedures for constructing and conducting photo arrays and live lineups. For example, with photo arrays, it suggests several unusual practices, among them (1) showing only one suspect per identification procedure, (2) selecting photos of "fillers"—i.e., nonsuspects—that match the eyewitness’s description of the criminal rather than the person the investigators suspect in the crime, (3) avoiding the use of nonsuspects who so closely resemble the suspect that a person familiar with the suspect might have difficulty distinguishing between the nonsuspect and the suspect, and (4) placing a suspect in different positions—i.e., reordering them—in each photo array when dealing with more than one eyewitness in a given case. ( Id. pp. 29–30.) Each of these procedures reduces eyewitnesses’ chances of misidentification.

The weight a good identification deserves

Although the Guide’s focus was on preventing eyewitness error rather than on catching errors after they had occurred, the Guide’s drafters recognized that the institution of new procedures will inevitably alter what goes on in the courtroom. In fact, during the drafting of the Guide, law enforcement personnel were hesitant about the project because of their fear that, in the interim between announcing the procedures and the mobilization of resources to put those procedures into routine practice, defense lawyers would use the new, ideal procedures to attack police still using the less optimal ones and thereby free guilty persons. After all, prosecutors argued, many identifications are so robust that suboptimal procedures do not degrade them.

Although prosecutors worried that the Guide’s procedures might be used as a stick with which to beat noncomplying investigators, they also recognized that the Guide’s procedures offered a carrot, too: a powerful persuasive weapon for investigators who did follow the Guide. Adherence to the Guide’s procedures, when properly understood and presented by trial prosecutors, virtually ensures that carefully collected eyewitness evidence will be given the weight that it deserves.

In the question-and-answer examples presented below, our assumption is that the police did follow a number of the new research-identified features outlined in the Guide for the composition and administration of live lineups. As such, these hypotheticals highlight significant departures from practices we have seen in the past. First, the "fillers"—i.e., the nonsuspects—were chosen to match the eyewitness’s description of the suspect rather than the features of a person the police may have apprehended as a suspect in the case. Second, the witness was properly instructed that the actual suspect might not even be present in the lineup. Third, the suspects were displayed one at a time—sequentially—rather than in a line at the same time, thereby avoiding the possibility that an eyewitness might choose a suspect based on the fact that he looked most like what the eyewitness remembered rather than on a genuine recognition of the suspect. Finally, in these examples, the lineup was administered "blind"—that is, using the technique of assigning the lineup procedure to an officer who did not know which lineup member was the suspect. Although this technique was not officially adopted in the Guide, it is identified there as a direction for future exploration and field testing. "Blind" assignment is a technique that has been adopted in the New Jersey Attorney General’s Guidelines for Preparing and Conducting Photo and Line Lineup Identification Procedures, the first set of state eyewitness evidence procedures based on the recommendations in the Guide.

Notice how the prosecutor in these examples mobilizes the careful use of these techniques to create a powerful direct examination and to preempt anticipated defense attacks:

Q. Can you describe a lineup for us?

A. A lineup is a procedure in which a suspect is displayed to an eyewitness along with others who meet the general description given by the witness. The idea is to see whether the witness is able to distinguish, and identify, the person involved in the crime.

Q. Is it important to conduct those procedures in a particular fashion?

A. Yes, our goal is to conduct the procedure in a manner that is fair for everyone involved, both the witness and the suspect.

The prosecutor elicits further information on the suspect "match-to-description" lineup composition technique:

Q. So, how do you do that—make it fair for everyone involved?

A. Well, it involves a variety of steps, including the selection of people to be displayed to the witness, instructions to the witness beforehand, and the way in which the people are displayed to the witness. Those sort of things.

Q. Tell me about the selection of people to display to the witness. How did you choose those persons?

A. I selected people that matched the original description given by the witness. In this case, the description referenced a white male, 6 feet tall,190 to 200 pounds, with brown hair and a mustache. So I selected men who met that description.

Q. Well, that description could fit a lot of people, some of who might look quite different from one another.

A. That’s true. But it is not a problem.

Q. Why is that?

A. You might think that you’d want people to all look just like the suspect—that’s what I used to do. But the problem is that we can’t be sure that our suspect is the person who committed the crime, at least until after the lineup. So to choose people that all look like the suspect is really not fair, to the suspect or the witness.

Q. How so?

A. Well, to the extent that police select people that all look like the suspect, you have narrowed the universe of options for the witness. This risks creating a subtle suggestion to the witness about what the police think the suspect looks like.

The prosecutor probes further regarding lineup instructions for witnesses:

Q. What about instructions? What do you tell the witness about the lineup in advance?

A. We told him a bit about what to expect and about how it would work. But there are several other specific instructions that are important.

Q. Like what?

A. I like to tell the witness that a lineup is important not just to identify the guilty, but to clear the innocent as well. I just don’t want them to feel pressure to make an identification.

Q. Do you say anything else about this?

A. I specifically tell the witness that the person who committed the crime may or may not be in the lineup.

Q. Isn’t that obvious?

A. Perhaps, but some witnesses could infer too much from the fact that the police have a suspect and are conducting the lineup.

Q. Why would that be a problem?

A. Well, I think that could lead to some witnesses selecting the person who looks most like the one that committed the crime with the idea that this will help move the investigations along, even if they are not certain. I want to guard against that.

Q. Does that make people overly cautious?

A. Perhaps, but I want to be able to rely on any identification that is made without fear that the witness was either unsure or trying too hard to help.

Sequential presentations in a live lineup are covered next:

Q. Tell us how you conducted the lineup after the witness was instructed.

A. I choose to present the lineup members to the witness one at a time, in sequence.

Q. Describe how that was done.

A. After the witness was seated in the darkened auditorium, we had each member of the lineup go into the lit lineup area one at a time. They each performed the same tasks—walk, turn, and face the witness—and then returned to the staging area before the next person went out.

Q. Why not show them all at once?

A. When you show them all at once, there is a greater chance that the witness will simply pick the person who looks most like the suspect, sort of a process of elimination.

Q. Why would that be a problem?

A. Well, it’s not, as long as you can be sure that the actual criminal is in the lineup. I mean, no one can look more like the criminal than the criminal himself, so there is no risk of a misidentification.

Q. So what’s the problem?

A. Well, most of the time, I may have an idea about who committed the crime, but I wasn’t there.

Q. So how does this one-at-a-time procedure address this?

A. By viewing the lineup members separately, the witness is in a better position to view each person and to make a comparison to their own memory of the crime. In other words, there is less opportunity to use a process of elimination because there is less of an opportunity to compare the lineup members.

Q. Describe how it was set up when the witness viewed the lineup.

A. We had her sit in the darkened auditorium with an officer. The officer read instructions and helped her fill out the identification sheet after the lineup was conducted.

Q. At the conclusion of this procedure, was the witness able to make an identification?

A. Yes. She chose Number 4.

Q. And who was Number 4?

A. The defendant.

Q. So the witness selected the person who you had arrested?

A. Yes.

"Blind" testing lineups are also explored:

Q. Is it possible that the officer influenced the witness to select the defendant, not intentionally, but unconsciously, perhaps?

A. No, that could not happen, not the way we run the lineup.

Q. How can you be sure?

A. Well, it’s impossible because the officer with the witness did not know who our suspect was. She came over from another unit to work with the witness and had nothing to do with the investigation. Our procedure is to ensure that the officer working with the witness has no information to distinguish the suspect from the stand-ins.

The price of shortcuts

Defense lawyers who become familiar with the Guide and its procedures will add an important new weapon to their arsenals, a weapon that they can use when the improved procedures are not followed. How this new tool can be employed must remain an open question. Are there courts that will allow the Guide to be used directly, on analogy to "learned treatises"? Perhaps. Even if direct use of the Guide is delayed or denied altogether, however, the procedures advocated therein are likely to find their way into local police field manuals, a traditional cross-examination tool. Besides, even if the Guide is never mentioned explicitly, defense lawyers who educate themselves in what could have, or should have, been done to protect eyewitness memory traces will enhance their ability to contrast for the jurors the inferior procedures used in this case. ( See generally Elizabeth Loftus & James Doyle, Eyewitness Testimony: Civil and Criminal ch. 10 (Lexis Law Books, 3d ed. 1997).)

The examples provided below emphasize points in the investigation at which the police failed to employ the Guide’s procedures. In an effective defense presentation, these themes would be foreshadowed in opening statement, explained in arguments, and form the backbone of an integrated theme, namely, "they could have done it right; they did it wrong; you jurors are entitled to doubt this evidence."

The defense attorney investigates a failure to compose a "match-to-description" photo lineup:

Q. In his picture, and today, Jon has shoulder length hair?

A. Yes.

Q. The five other men in the photo spread had hair at or below their shoulders?

A. Yes.

Q. But no one in the photos had short hair, like Maude described on the person she saw driving the car?

A. No. I wanted all the photos to look like Jon so he wouldn’t stand out, to avoid being suggestive.

Q. So, no one in the photo spread looked like the man she actually saw?

A. Well, no. I just didn’t want her to be confused. They were all white men around 30 years old.

Q. You didn’t put a couple extra photos in the spread of men with short hair just in case?

He continues regarding the failure to use sequential rather than simultaneous presentation:

Q. Your procedure for how you showed Maude the pictures was that you laid them all down on the table at the same time?

A. Yes.

Q. You never showed her each photo individually and asked if that was the driver?

A. No. We never do it that way.

Q. You testified that you told her to take her time and look at all of them?

A. Yes.

Q. And she did?

A. Yes.

Q. This would allow her to pick out one she thought looked most like the person she saw driving the car?

A. Well, I told her to see if she saw the person who was driving the car in any of the pictures.

Q. Every time she saw . . . she had others to compare it to?

A. Yes.

Q. Always multiple choices, never yes/no?

The absence of prelineup instructions to witnesses is also mentioned:

Q. You didn’t tell her that this lineup was as important to clear innocent people as it was to catch guilty ones?

A. No.

Q. You didn’t tell her that the driver’s picture might not be there, did you?

A. No.

Q. You didn’t assure her the investigation would continue whether or not she was able to make an identification of anyone, the procedure from the Department of Justice Guide?

A. No. We don’t usually say that.

Failure to keep witnesses uncontaminated is also brought out:

Q. She especially knew about other witnesses when she signed the back of Jon’s photo and saw her neighbor Ann’s signature there, didn’t she?

A. Well, she didn’t see that before she picked the photo, so it didn’t matter.

Q. Before she saw that signature, she said she was "pretty sure" it was Number 4 [Jon’s photo] who was the driver but that "the guy I saw looked like he had short hair"?

A. Yeah.

Q. Ann is the mother of the boy who was shot?

A. Yes.

Q. After Maude saw Ann’s signature and signed the photo herself, you told her "These are old photos. We’ll call you when we’re ready for a lineup"?

A. That’s right.

Q. [Reading and then referring to the Guide.] Your procedures do not include instructing witnesses not to discuss the details of the events with each other?

A. Not exactly. We don’t let them look at lineups together, though.

Q. But your photo lineup procedure included giving an eyewitness information of another identification?

A. Well, I didn’t give her any information; she just saw the signature after she made her identification.

Q. But the procedure wasn’t finished was it?

A. What do you mean?

Q. You still wanted her to see a lineup; there was further identification evidence to be received from her?

The procedures used in conducting a physical lineup impact the reliability of the identification evidence, as investigated here:

Q. I want to talk about your procedures for the lineup that occurred at the police station. Four days after the photo spread, you called Maude and said, "We’re ready to have you see a lineup at the station"?

A. Yes.

Q. When she arrived, you told her that you had caught a man driving the green car that was used in the shooting?

A. I don’t remember telling her that, but it’s possible. I think I told her that after the lineups were over. I’m not sure.

Q. You set up the lineup?

A. Yes.

Q. Jon was the only one from the photo spread who was also in the lineup?

A. Yes.

Q. Jon’s hair was at his shoulders in the lineup, and he had pock marks on his face?

A. Yes.

Q. Maude said the driver had short hair and didn’t see any marks on his face?

A. Yes.

Q. The other men in the lineup had longish hair, at least shoulder length, and clear skin?

A. Yes, that was the best we could do.

Q. You didn’t ask them to pull their hair back?

A. No.

Q. As Maude viewed the lineup, you asked her if she "recognized anyone"?

A. From the shooting, yes.

The absence of live prelineup instructions is also noted:

Q. Again, you didn’t tell Maude that the person who drove the car at the time of the shooting might not be in the lineup?

A. No.

Q. That it was just as important to clear the innocent as to identify the guilty?

A. No.

Q. That the investigation would continue whether or not she was able to make an identification of anyone?

A. No.

Q. After she picked Jon, you never asked her to tell you, in her own words, how sure she was about her identification of him?

A. No . . . she picked Jon . . . that was it. I thanked her, and she went home.

Q. And when she came back before testifying, you showed her the same picture?

What’s not in the Guide

The Guide’s recommendations stopped short of anything that researchers would consider state-of-the-art. The Guide does not, for example, recommend "blind" identification procedures. While it includes sequential display of photographs and live suspects, it does not endorse the researchers’ strong preference for that method. Nor does the Guide require videotaping of procedures. However, the Guide never set out to be a compendium of cutting-edge research. Rather, the Guide was a collaboration that sought to combine the best practices from the law enforcement community tempered by research findings that could be readily adapted for present-day police use. In this exchange, police were able to educate researchers about the practical issues that they face in the field, while researchers educated law enforcement about their findings. The results of this exchange appear to have been productive. For example, after researchers learned about law enforcement’s concern with running sequential photo arrays in the field, work was begun to design a computer program that could resolve the issue.

What’s next?

The Guide is significant only as a starting point. After the long freeze in the relationship between law enforcement and psychological researchers, the NIJ’s Technical Working Group provided a meaningful forum in which to exchange ideas and wrestle with common concerns about eyewitness identification. The publication of the Guide is neither a great triumph for researchers nor a threat to principled, hardworking police officers. The truth is that the world was not anxiously awaiting the next publication from the federal government instructing local government on how to order its affairs. The hope for the future lies not so much not in what was published but in the continued collaboration that the effort launched.

This collaboration can continue to germinate if the parties involved resist the urge to bicker about who wins or loses. Researchers can accept the fact that they did not get all that they wanted. Law enforcement personnel can remain open to the possibility that their day-to-day activities can be informed and improved by academic research. Thankfully, there are models to follow as we take our next steps forward.

Dr. John Turtle and Dr. Rod Lindsay, both researcher members of the Technical Working Group, have crossed the divide and developed excellent relationships with law enforcement. Veteran police investigators Paul Carroll, Ed Rusticus, and Ken Patenaude, also members of the group, have embarked on an ambitious program of talking (and listening) to police about the Guide’s findings. The New Jersey Office of the Attorney General has integrated many of the Guide’s findings—as well as additional research findings not embodied in the Guide—into that state’s practice. Ideas and training are currently being exchanged freely, with none of the acrimony that often pervades the debate surrounding cases headed for court. These efforts must be nurtured to ensure that new ideas can be developed. The cutting edge of this field will not be felt in laboratory demonstrations but when working cops apply them in the field.

The NIJ has remained involved and developed training materials for law enforcement based on the Guide. These are expected to be distributed shortly to every police department in the country. The NIJ must also develop a plan to have that training delivered to its target audience. There is plenty of work left for everyone. The existence of training materials will be of little value without a strategic plan for implementation. Lawyers on both sides of the system will have to educate themselves—and educate the judges, too. The worst of all worlds would be to have created standards for eyewitness evidence that were never effectively communicated to the practitioners of that science—individual police officers.

Life can be better for everyone—except, of course, the guilty—if we continue to work to ensure real progress in the "friendly and energetic alliance" envisioned by Wigmore. Hopefully, the NIJ will continue to foster this exchange to finish what it has nobly begun through the publication of the Guide.


James M. Doyle is of counsel to the Boston firm of Carney & Bassil. The former head of the statewide Public Defender Division of the Massachusetts Committee for Public Counsel Services, he is the coauthor of Eyewitness Testimony: Civil and Criminal . Mark R. Larson has been the King County’s chief criminal deputy since 1993 in the state’s attorney’s office in Seattle, Washington. He is also a member of the faculty of the National Advocacy Center. Caterina M. DiTraglia is an assistant federal defender in St. Louis, Missouri. She is the former head of the Missouri State Public Defenders Youth Advocacy Project and lectures frequently on capital defense, juvenile process issues, and trial advocacy. All three authors were members of the NIJ’s Technical Working Group for Eyewitness Evidence.

Return to Table of Contents - Fall 2001

Return to Criminal Justice magazine home page